A trust game model for the cognitive radio networks

Security in cognitive radio networks is still an open challenge especially with the increased number of attacks. Many approaches were proposed in literature in order to secure the CRNs such as the trust management and reputation. However, most of existing trust approaches focus on the attacks detection and do not take into consideration the penalty of the malicious users. To address this issue, we introduce a trust game model for the cooperative sensing spectrum in cognitive radio networks in order to encourage the secondary users to choose the honest strategy by sending correct sensing outcomes. The main feature of this model is the punishment function aimed at exempting the malicious users with faulty spectrum sensing from participating in the spectrum detection stage during a predefined period. Our proposal switches periodically between the first-hand and second-hand sensing evidences to improve the network performance.